Today, Google released two new generative AI models designed to enable developers to experiment, iterate on, and scale their ideas faster and at lower cost. As the latest member of the Nano Banana family—the Gemma image model lineage—Gemini 3.1 Flash Lite Image (Model ID: gemini-3.1-flash-lite-image) is engineered for scenarios requiring high throughput and ultra-low latency. It generates images from text prompts in just four seconds, with costs as low as $0.034 per 1K-resolution image. This represents a significant speed increase without compromising prompt adherence, character consistency, or the quality of text rendering within images. Official guidance recommends that developers previously using Gemini 2.5 Flash Image switch immediately to this new model to achieve comprehensive performance improvements. The model has been made available via Google AI Studio, the Gemini API, and the Gemini Enterprise Agent Platform, while gradually rolling out across consumer-facing products such as AI Mode in Search, the Gemini app, NotebookLM, and Google Photos. Additionally, Gemini Omni Flash, which debuted earlier today at Google I/O, is now officially open to developers (Model ID: gemini-omni-flash-preview). Priced at $0.10 per second, it competes directly with Veo 3.1 Fast. The natively supports high-quality video generation and conversational editing through language based on inputs comprising text, images, and videos. Leveraging Gemini’s multimodal reasoning capabilities, it constructs logically coherent video content tailored to specific contexts such as history, biology, and narrative logic. Currently, each output can be up to 10 seconds long, with longer durations coming soon. Audio reference uploads and scene extension features have yet to become available via the API. By utilizing Gemini 3.1 Flash Lite Image for rapid image generation and then passing those generated images as references into Omni Flash for animation, users can create complete end-to-end multimodal experiences. Alongside these releases, Google introduced three demonstration applications: Anywhere (virtual travel), Space Lift (interior design), and Omni Product Studio (e-commerce video production). Both models incorporate SynthID digital watermarking to ensure content transparency and regulatory compliance.